Development, Freedom, and Happiness: A Global Perspective

Transcription

1 Development, Freedom, Happiness Perspectives on Psychological Science 1 Development, Freedom, and Happiness: A Global Perspective Ronald F. Inglehart Department of Political Science University of Michigan (Ann Arbor, MI USA) Roberto Foa Department of Government Harvard University (Cambridge, MA USA) Christopher Peterson Department of Psychology University of Michigan (Ann Arbor, MI USA) Christian Welzel School of Humanities and Social Sciences Jacobs University (Bremen, Germany) Forthcoming in Perspectives on Psychological Science (July, 2008), published by Blackwell Publishing for the Association for Psychological Science. The definitive version is available at The authors gratefully acknowledge support from the National Science Foundation, and the foreign ministries of Sweden and The Netherlands, each of which supported fieldwork in several countries in the wave of the World Values Survey. Address all correspondence to Ronald Inglehart, Institute for Social Research, University of Michigan, Ann Arbor, Michigan Telephone: , FAX: ; RInglehart@gmail.com

2 Development, Freedom, Happiness Perspectives on Psychological Science 2 Abstract Until recently, it was widely held that happiness fluctuates around set-points, so that neither individuals nor societies can lastingly increase their happiness. Even though recent research showed that some individuals move enduringly above or below their set-points, this does not refute the idea that the happiness levels of entire societies remain fixed. Our article, however, challenges this idea: Data from representative national surveys carried out from 1981 to 2007 show that happiness rose in 45 of the 52 countries for which substantial time-series data were available. Regression analyses suggest that that the extent to which a society allows free choice has a major impact on happiness. Since 1981, economic development, democratization, and rising social tolerance have increased the extent to which people perceive that they have free choice, which in turn has led to higher levels of happiness around the world, as the human development model suggests. KEY WORDS: free choice, happiness, human development, life satisfaction, set-point, subjective well-being, tolerance

3 Development, Freedom, Happiness Perspectives on Psychological Science 3 Development, Freedom, and Happiness: A Global Perspective 1. INTRODUCTION Psychologists, economists, biologists, sociologists, and political scientists have long investigated human happiness, and until recently one claim found widespread acceptance: Happiness remains constant. Research implies that neither rising prosperity nor severe misfortune permanently affect happiness. After a period of adjustment, individuals return to their baseline levels of well-being, leaving humanity on a hedonic treadmill (Brickman & Campbell, 1981; Diener, Suh, Lucas & Smith, 1999; Kahneman, Krueger, Schkade, Schwartz, & Stone, 2004). Similarly, as entire countries become richer, relative gains and losses neutralize each other across populations, bringing no overall increase in the happiness of their citizens (Easterlin, 1974; Kenny, 2004). Moreover, biological factors are closely linked with a sense of well-being (Ebstein, Novick, Umansky, Priel, & Osher., 1996; Hamer, 1996), and twin studies suggest that happiness is heritable to a significant extent (Lykken & Tellegen, 1996; Lyubomirsky, Sheldon, & Schkade 2005). Individual differences in happiness may therefore reflect inherent differences in temperament and consequently resist change (Diener & Lucas, 1999). A widely-accepted view is that happiness fluctuates around a fixed set-point (Headey & Wearing, 1989; Larsen, 2000; Williams & Thompson, 1993). Insofar as this setpoint is biologically determined, neither individual efforts nor social policy can bring lasting changes in happiness. Another explanation for the apparent stability of the aggregate happiness of nations is social comparison theory (Easterlin, 1974, 2003). According to this account, happiness stays the same in the face of rising income because of a shift in reference. If happiness is shaped by one s relative position in a society, then even if a nation s overall economy grows, only those with above-average gains will experience rising happiness, and these increases will be offset by decreases among those with belowaverage gains. A large body of evidence does indicate that the subjective well-being levels of given countries are stable. For example, Inglehart (1990) demonstrated that life satisfaction levels were very stable from 1973

4 Development, Freedom, Happiness Perspectives on Psychological Science 4 to 1988 in most West European countries. Diener and Oishi (2000), Inglehart and Klingemann (2000), Easterlin (2005) and Kahneman and Krueger (2006) presented similar findings. But the strongest support for the claim that the happiness levels of countries have not risen over time comes from the United States, which provides by far the longest and most detailed time series data on subjective well-being. Hundreds of surveys have measured happiness and life satisfaction among the American public in almost every year since No other country has a comparable database, and the US data show a flat trend from 1946 to the present. Because the happiness levels of given societies do not seem to change over time, the idea that economic development brings rising happiness has been widely rejected. Although rich nations show higher levels of subjective well-being than do poor countries, these differences may reflect cultural differences in what happiness means. THEORETICAL DISCUSSION Can Happiness Change? Recent research questions the view that the happiness of individuals is unchanging. Diener, Lucas and Scollon (2006) demonstrated that the subjective well-being levels of some people can and do change over time. Fujita and Diener (2005) analyzed data from a 17-year German panel study, finding that 24% of the respondents life satisfaction levels changed significantly from the first five years of the study to the last five years. Similarly, Lucas, Clark, Georgellis, and Diener (2005) analyzed a 15-year longitudinal study of the effects of marital transitions on life satisfaction. They found that on average, individuals moved back toward their baseline levels of satisfaction but that a significant numbers of individuals remained above their original baseline level, while other remained below it. Individuals are not necessarily trapped on a hedonic treadmill. What about nations? Findings that happiness can change for individuals do not necessarily mean that the happiness levels of given societies change. Social comparison theory holds that the relative gains and losses of different individuals in a given nation will cancel each other out, resulting in no discernible shifts, upward or downward, for a society as a whole.

5 Development, Freedom, Happiness Perspectives on Psychological Science 5 Cross-sectional comparisons of nations show considerable variation in the happiness of their people and that economic development is strongly correlated with happiness: Inglehart (1990:chapter 1) analyzed data from 24 countries covering a wide range of economic levels and found a.67 correlation between per capita GNP and life satisfaction. He interpreted this as implying that economic development is conducive to rising happiness. But up to the present time, longitudinal evidence has not provided compelling support for this interpretation. Hagerty and Veenhoven (2003) argued that growth does increase happiness, demonstrating that income was positively correlated with happiness in 14 of the 21 nations for which data were available from 1972 to However, Easterlin (2005) argued that their findings were not robust at the level of specific nations and relied on different measures of happiness administered to different types of samples and failed to control for seasonal effects. And even though they demonstrated an increase in happy life years across all nations for which considerable time series data are available, this result is almost entirely driven by the increase in life expectancy and not by an increase in happiness itself. More recently, Hagerty & Veenhoven (2006) demonstrated statistically significant increases in subjective well-being in four of eight high-income countries, and three of four low-income countries for which a long time series was available, but the evidence did not seem decisive. The authors summed up the debate as follows: The difference arises from the fact that the available data are not too clear and therefore allow different interpretations. Easterlin reads the data as showing that the glass is half empty, while we see the glass to be half full. (Hagerty & Veenhoven, 2006:p.422). The claim that happiness has risen in most countries remains unsettled, and is not generally accepted. The present paper reports cross-national longitudinal findings that are not undermined by problems of methodological incomparability. Data from the Values Surveys asked the same questions in the same format to large representative national samples of respondents in scores of countries from 1981 to 2007, allowing a more definitive test of the hypothesis that happiness has increased in most countries. As we will see, it has. Theoretical Frame: Human Development and Happiness

6 Development, Freedom, Happiness Perspectives on Psychological Science 6 As important as it is to determine if the happiness of nations has changed, it is even more important to understand why it may have changed. Economic development within a nation is a likely starting point for any explanation because it is demonstrably associated with psychological changes that in turn should impact people s happiness (Diener, Diener & Diener, 1995). Inglehart (1997) hypothesized that economic development brings a societal-level shift from maximizing economic growth to maximizing subjective well-being: The transition from a society of scarcity to a society of security brings a dramatic increase in subjective well-being. But we find a threshold at which economic growth no longer seems to increase subjective well-being significantly. This may be linked with the fact that at this level, starvation is no longer a real concern for most people. Survival begins to be taken for granted. Significant numbers of Postmaterialists begin to emerge and for them, further economic gains no longer produce an increase in subjective wellbeing. From a rational actor's perspective, one would expect economic development to eventually bring a shift in survival strategies. Figure 2-4 suggests how this works. At low levels of economic development, even modest economic gains bring a high return in terms of caloric intake, clothing, shelter, medical care and ultimately, in life expectancy itself. For individuals to give top priority to maximizing economic gains, and for a society to give top priority to economic growth, is a highly effective survival strategy. But once a society has reached a certain threshold of development one reaches a point at which further economic growth brings only minimal gains in both life expectancy and in subjective well being. There is still a good deal of cross-national variation, but from this point on non-economic aspects of life become increasingly important influences on how long, and how well, people live. Beyond this point, a rational strategy would be to place increasing emphasis on quality of life concerns, rather than to continue the inflexible pursuit of economic growth as if it were a good in itself (Inglehart, 1997, pp ).

7 Development, Freedom, Happiness Perspectives on Psychological Science 7 Insert Figure 1 about here [Figure 2-4 from Inglehart, 1997] This societal-level shift is linked with individual-level value changes, from giving top priority to economic and physical security toward giving top priority to self-expression values, which emphasize participation, freedom of expression, and quality of life. Under conditions of scarcity, people focus on survival needs, giving top priority to economic and physical security. Economic development increases people s sense of existential security, leading them to shift their emphasis from survival values toward self-expression values and free choice, which is a more direct way to maximize happiness and life satisfaction. The underlying theme of this shift in life strategies is to deemphasize external authority and maximize individual autonomy. How it operates is spelled out more fully in the human development model proposed by Welzel, Inglehart & Klingemann (2003) and Inglehart & Welzel (2005). Human development shifts emphasis from the pursuit of happiness through economic means, toward a broader pursuit of happiness by maximizing free choice in all realms of life. The belief that one has free choice and control over one s life is closely linked with happiness (Johnson & Krueger, 2006), and this link seems to be universal. Happiness is linked with people s sense of freedom across all major cultural zones (Inglehart & Welzel, 2005:140). In many societies, people value free choice as much as they value economic security (Sen, 2001) and emphasis on freedom increases with rising economic security. The fact that people change the way in which they pursue happiness, does not necessarily mean that they will attain it. But since 1981, these shifts in individual-level values have contributed to societal changes that are conducive to human happiness. Since 1981, self-expression values have become increasingly widespread around the world, contributing to democratization, growing support for gender equality, and growing acceptance of outgroups such as gays and lesbians (Inglehart & Welzel, 2005). Inglehart (1990) found that happiness is strongly related to democracy (Barro, 1999; Frey & Stutzer, 2000; Inglehart & Klingemann, 2000). This relationship seems to be reciprocal: High levels of

8 Development, Freedom, Happiness Perspectives on Psychological Science 8 subjective well-being are conducive to democracy, and democracy provides a wider range of free choice, which is conducive to subjective well-being (Haller & Hadler, 2004; Inglehart & Welzel, 2005; Ott, 2001; Veenhoven, 2000; Welsch, 2003). Like democratization, social tolerance broadens the range of choices available to people, enhancing happiness. Accordingly, Inglehart & Welzel (2005) found that support for gender equality and tolerance of outgroups were strongly linked with happiness not just because tolerant people are happier, but because living in a tolerant society enhances everyone s freedom of choice. Similarly, Schyns (1998) argued that gender equality is linked with happiness. During the late 1980s and early 1990s, dozens of societies experienced transitions to democracy that enhanced freedom of expression, freedom to travel, and free choice in politics. Moreover, from 1981 to 2007, support for both gender equality and tolerance of outgroups increased substantially in most of the countries monitored by the Values Surveys (Inglehart & Welzel, 2005). Furthermore, during the past two decades, low-income countries containing fully half of the world s population have experienced one of the highest rates of economic growth in history, allowing them to emerge from subsistence-level poverty. By a favorable combination of circumstances, societal changes of the past two decades have increased the prosperity of people in less-prosperous societies, and the political and social freedom of people in middleincome and high-income societies, enhancing the extent to which people in both types of societies have free choice in how to live their lives. We hypothesize that these changes have been conducive to rising levels of happiness within entire societies. THE PRESENT STUDY The happiness of nations was investigated with data from the World Values Survey and European Values Study, which have carried out five waves of surveys from 1981 to 2007 in scores of countries containing almost 90% of the world s population. All five waves of surveys included two widely-used indicators of subjective well-being (a) happiness and (b) overall life satisfaction administered in the same format in equivalent translations in every wave.

9 Development, Freedom, Happiness Perspectives on Psychological Science 9 To test whether happiness levels have risen, we constructed a subjective well-being (SWB) index using these two indicators. The index provides a broader-based and more reliable indicator of the subjective well-being levels of given societies than either of its two components. We examined the trends on this indicator and its two components in 52 societies. To test the hypothesis that factors reflecting free choice were conducive to rising levels of happiness, we carried out OLS panel regression analysis to analyze societal-level effects, and HLM regression analysis to test the interaction of individual-level and societal level effects, using the SWB index, happiness, and life satisfaction as dependent variables, and measures of the feeling that one has free choice, and related attitudes as independent variables controlling for democratization and growth in GDP/capita. 1. METHOD Research Participants and Procedure Over the past few decades, the Values Surveys have interviewed representative national samples of scores of countries, with an average sample size of 1,400 respondents. Over a thousand publications have been based on these data. Extensive information about publications, findings, fieldwork and the organization of these surveys can be found at 1. The present article presents the results of cross-sectional analysis based on data from scores of nations containing almost 90 percent of the world s population. The longitudinal analyses presented here are based on data from the 52 nations for which substantial time series data are available. Substantial is defined as having surveys from at least two waves that were scheduled to be held at least ten years apart. Because actual fieldwork sometimes took place earlier or later than targeted, in some cases the actual time span between surveys was less than ten years, and in other cases the span covered as much as 26 years. For the average country, we analyzed the changes that took place over a period of 17 years, as measured by almost 4 surveys per country. Appendix A shows the year of fieldwork for each of the 194 surveys, and the levels of happiness, life satisfaction, and the SWB index score from that survey. Measures

10 Development, Freedom, Happiness Perspectives on Psychological Science 10 Life satisfaction was assessed by asking respondents to indicate how satisfied they were with their life as a whole, using a scale that ranged from 1 [not at all satisfied] to 10 [very satisfied]. Happiness was assessed by asking respondents to indicate how happy they were, using four categories: very happy; rather happy; not very happy; and not at all happy. These items are sensitive indicators of a broad subjective well-being dimension (Andrews & Withey, 1976), capturing most of the common variance in scores of domain-specific indicators. They were included in all five waves of the Values Surveys. For a composite measure of subjective well-being, we combined each person s responses to the questions about happiness and life satisfaction to produce an SWB index, giving equal weight to each variable. Because life satisfaction is measured on a ten point scale, and happiness on a four-point scale, and because the two questions have opposite polarity, the SWB composite was constructed as follows: SWB = life satisfaction 2.5 * happiness. If 100 % of its people were very happy and extremely satisfied, a country would get the maximum score of 7.5. If happiness and life satisfaction were evenly balanced, the country would get a score of zero. If more people were dissatisfied/unhappy than are satisfied/happy, the country would get a negative score. Respondents also indicated to what extent they felt they had free choice and control over their lives, using a scale that ranged from 1 [none at all] to 10 [a great deal]. They also designated groups of people that they would not like to have as neighbors, using a list that included drug addicts, people of a different race, people with AIDS, immigrants/foreign workers, homosexuals, people of a different religion, and heavy drinkers. Responses to these items correlated strongly, with acceptance or rejection of homosexuals being a particularly sensitive indicator of overall tolerance of outgroups. Respondents also indicated to what extent they felt that homosexuality can be justified, using a scale that ranged from 1 [never] to 10 [always]. A number of items in these surveys tapped religiosity, but the most sensitive indicator asked: How important is God in your life? using a 10-pont scale ranging from 1 [not at all important] to 10 [very important]. Respondents were also asked, How proud are you to be [COUNTRY S NATIONALITY]? Response options were very proud, quite proud, not very proud and

11 Development, Freedom, Happiness Perspectives on Psychological Science 11 not at all proud. These items were used to test, respectively, the impact of a sense of free choice, tolerance of outgroups, religiosity and national pride on subjective well being. To measure the impact of economic factors and democratization, we use the society s GDP/capita (purchasing power parity estimates) and economic growth rate from the World Bank database; and a measure of democracy from the Polity IV project (for information about this project, see 3. RESULTS Overall, here is what we found. Cross-sectional data from the 1990 Values Surveys suggested that economic development is conducive to rising levels of subjective well-being, and Inglehart (1997) and Hagerty & Veenhoven (2003) proposed that interpretation. Longitudinal evidence did not seem to support it, however, and it was not generally accepted. The following analyses test the hypothesis that factors conducive to human choice, such as economic development and democratization, do bring rising levels of subjective well-being, using a broader and more reliable longitudinal data base than was previously available. Cross-sectional analysis of recent data confirms that economic development is indeed strongly linked with high levels of subjective well-being, but also indicates that economic factors are only part of the story. Pooled time series regression analysis suggests that religion, tolerance of outgroups, and a society s level of democracy are strong predictors of subsequent levels of subjective well-being, controlling for economic development and a society s initial level of subjective well-being. Analysis of changes over time suggest that all of the foregoing factors influence subjective wellbeing mainly in so far as they enable people to have a wider range of free choice. Since the years since 1981 have been a period of global economic growth, widespread democratization and rising social tolerance, this implies that subjective well-being should have increased. And longitudinal evidence from 52 countries for which substantial time series data are available, demonstrate that from 1981 to 2007, feelings of free choice did increase and subjective well-being increased in an overwhelming majority (fully 77 percent) of these countries.

12 Development, Freedom, Happiness Perspectives on Psychological Science 12 Economic Development and Happiness Figure 2 shows the relationship between economic development and subjective well-being in 88 countries containing almost 90% of the world s population. To maximize reliability, this figure is based on data from all surveys carried out in a given country from 1995 to We used per capita GDP five years before a country s mean survey, to reflect the time lag between reaching a given level of prosperity and its impact on subjective well-being. The solid curve on Figure 2 depicts the cubic regression line for the relationship between per capita GDP and subjective well-being. If a society s subjective well-being were wholly determined by its level of economic development, it would fall on this line. Happiness and life satisfaction rise steeply as one moves from subsistence-level poverty to a modest level of economic security, and then level off. 2 Among the richest societies, further increases in income are only weakly linked with higher levels of subjective well-being. Insert Figure 2 about here As Figure 2 indicates, the aggregate subjective well-being of nations is closely related to economic development (r =.62, p <.001). The people of high-income countries are much happier and satisfied with life than are the people of low-income countries, and the differences are substantial. In Denmark, 52% of the public indicated that they were highly satisfied with their lives (placing themselves at 9 or 10 on a tenpoint scale), and 45% said they were very happy. In Armenia, only 5% were highly satisfied with their lives, and just 6% were very happy. In contrast to the small income-linked differences usually found within most countries, the cross-national differences are extremely large. The relationship between subjective well-being and economic development is curvilinear, explaining why it is so widely believed that economic development has no impact on subjective wellbeing. Previous longitudinal studies relied heavily on U. S. data, and the United States long ago made the transition from subsistence-level poverty to middle-income status. Among higher-income nations, further

13 Development, Freedom, Happiness Perspectives on Psychological Science 13 gains in income bring relatively little change in well-being (Frey & Stutzer, 2002; Inglehart, 1990, 1997; Myers, 1993; 2000). Figure 2 makes another important point. Quite apart from their level of economic development, some types of societies do a better job of maximizing their citizens subjective well-being than others. All twelve of the Latin American countries for which there were data showed higher levels of subjective wellbeing than their economic levels would predict. Conversely, although they are about as rich as Latin America, almost all of the ex-communist societies showed lower levels of subjective well-being than their economic levels would predict. Indeed, Russia and other ex-soviet states showed lower levels than much poorer countries such as India, Bangladesh, Nigeria, Mali, and Ethiopia. Life satisfaction and happiness showed similar patterns, with the Latin American societies being over-achievers and the ex-communist societies under-achievers on both indicators of subjective wellbeing. Across the Latin American countries, 45% of the population described themselves as very happy, and 42% rated themselves as very satisfied with their lives as a whole. In the ex-communist countries, only 12% described themselves as very happy, and only 14 % were very satisfied. Though their economic levels are fairly similar, Latin Americans were three to four times as likely to have high levels of subjective well-being. Communist rule is not necessarily linked with low levels of subjective well-being: China and Vietnam still ruled by communist parties and currently enjoying high rates of economic growth showed much higher levels of well-being than the Soviet successor states. The collapse of their political, economic, and belief systems seem to have sharply reduced subjective well-being in the ex-communist societies. Many of them have a damaged national myth. The Soviet Union and Yugoslavia once played prominent roles in the world, which may have brought feelings of pride and satisfaction to many of their citizens. Today, they have splintered into 21 diminished successor states. Although religion has long been weak in these countries, communist ideology may once have played a role comparable to that of religion. For many decades, communism seemed to be the wave of the future. The belief that they were building a better society may have given a sense of purpose to the lives

14 Development, Freedom, Happiness Perspectives on Psychological Science 14 of many people. It is difficult to understand the rise to power of the communist movements in Russia, China, and Vietnam without recognizing the motivating power of a belief system that once gave meaning to many people s lives and made them willing to sacrifice their lives for the cause. In Latin America, traditional beliefs in God and country remain strong; but in the ex-communist countries, the collapse of communism left a spiritual vacuum. The striking contrast between the subjective well-being levels of Latin America and the excommunist countries suggests that some types of society are more conducive to high levels of subjective well-being than others, quite apart from economic factors. Accordingly, dummy variables for Latin American and ex-communist countries explain a large share of the cross-national variance in SWB, even controlling for economic variables. But this does not explain why these countries show distinctive levels of SWB. The following regression analyses probed more deeply. To provide complementary perspectives on the factors contributing to changes in SWB, we first analyzed societal-level factors, using OLS panel regression with data from the 52 countries from which substantial time series data are available. Cases were weighted according to the length of time elapsed between surveys, so countries for which the full 26-year time series was available have 2.6 times the weight of countries with a ten-year series. We then examined interactions between societal-level factors and individual-level factors, using hierarchical linear modeling (HLM). The pooled time series regression results in Table 1 predicted the level of subjective well-being found in the latest survey from each country, controlling for the initial level of SWB. Models 1.1. and 1.2 use the SWB index as the dependent variable and Models 1.3 and 1.4 analyze changes in this index s two components, happiness and life satisfaction respectively. High levels of religiosity and tolerance of outgroups predicted relatively high future levels of subjective well-being. Especially under conditions of low economic security, religion provides a sense of predictability and security (Norris & Inglehart, 2004). Until recently, communist ideology filled this function for many people. The collapse of communist

15 Development, Freedom, Happiness Perspectives on Psychological Science 15 ideology, however, left a vacuum in many societies, contributing to a declining sense of well-being. To some extent, growing emphasis on religion has helped fill this void, with a significant impact on SWB. The extent to which people live in a tolerant society may also shape SWB. Intolerant social norms can narrowly restrict people s life choices, reducing subjective well-being. Tolerance of gender equality, gays and lesbians, people of other religions, foreigners and other groups, tends to be strongly correlated with each other; and at this point in history the most sensitive indicator of overall tolerance is tolerance of homosexuals, the least-liked group in most societies. This indicator has a significant impact on SWB. It is not just that being tolerant oneself makes one happy living in a tolerant social environment is conducive to happiness for everyone. 3 Alhough national pride had a strong zero-order correlation with SWB, it was closely linked with strong emphasis on religion, so when religiosity was included in the regression, national pride did not have an independent impact. Both religion and national pride were stronger in less developed societies than in developed ones, which helps explain why some low-income societies had relatively high levels of SWB. Thus, the contrast between the Latin American societies and the ex-communist societies shown in Figure 2 may be due in part to the fact that virtually all of the Latin American publics were strongly religious and had strong national pride, whereas the ex-communist nations did not: The proportion saying that God is very important in my life (placing themselves the top of a 10-point scale) was 76% in the Latin American countries and only 27% in the ex-communist countries. The figure in the remaining countries was 42%. The proportion saying that they are very proud of their nationality is 77% in the Latin American countries and only 39% in the ex-communist societies. The figure in the remaining countries was 57%. High levels of prosperity had a significant impact on subsequent levels of SWB and life satisfaction, but not on happiness. A society s level of life satisfaction seems more strongly influenced by economic conditions than is happiness. Although they experienced democratization, in most of the excommunist countries, the transition to democracy and a market economy was accompanied by severe economic decline. In Russia, for example, real income fell to less than half its pre-transition level, and life

16 Development, Freedom, Happiness Perspectives on Psychological Science 16 expectancy declined by several years. Consequently, in many of the ex-communist countries, happiness rose, but life satisfaction fell. High levels of democracy at the time of the first survey had a significant negative relationship with subsequent SWB. At first glance this seems counter-intuitive because the citizens of democracies tend to be happier than those of authoritarian societies: Our SWB index showed a.74 correlation with democracy as measured in 1987, just before the recent wave of democratization. The fact that a large number of unhappy societies suddenly shifted toward democracy reduced the correlation between SWB and democracy, because SWB is relatively stable. The correlation between SWB and democracy fell to about.4 by 1993 and remained at that level through But countries scoring low on SWB were much likelier to shift toward democracy than those that ranked high: The correlation between happiness and democratization (as measured by the amount of change in political rights and civil liberties scores) was Thus, countries that were democratic and happy at the start of the time series showed little or no change. Conversely, countries that initially ranked low on both subjective well-being and democracy were far more likely to shift toward democracy and thus were more likely to show increases in subjective well-being. Freedom of Choice and Rising Happiness Human development theory implies that the main reason why the changes of the past 25 years led to rising happiness is because they brought greater freedom of choice. Although our measure of the feeling of free choice had a strong zero-order correlation with SWB, it did not show a significant impact in Table 1 because its variance overlaps with other the variables. Nevertheless, growing freedom of choice had a major impact on rising happiness, as Table 2 demonstrates. This table analyzes changes in subjective well-being from the first available survey for each country to the most recent available survey. The results show that a growing feeling that one has free choice was by far the most important influence on whether SWB rose or fell. The feeling that one has free choice and control over one s life increased in many countries, with a powerful impact on changes in SWB. Indeed, when free choice was included in the regression, it was the only variable that showed a statistically significant association with change in

17 Development, Freedom, Happiness Perspectives on Psychological Science 17 SWB, regardless of whether the SWB index, happiness, or life satisfaction was the dependent variable. By itself, free choice explained 30% of the change over time in SWB. This analysis explained a smaller share of the change in happiness (21%) than in life satisfaction (31%) or the SWB index (30%). One explanation might be that while life satisfaction is measured on a 10-point scale, happiness is measured on a 4-point scale, which might produce less accuracy in measurement. Alternatively, a substantial share of the change in happiness is shaped by factors not identified here. In any case, the human development model seems to be on target in suggesting a sense of free choice as the vehicle to higher levels of SWB. Latin Americans ranked much higher here than the citizens of ex-communist countries: 45% of the former said they had a great deal of choice (points 9 or 10 on a 10-point scale) as compared with 21% among the latter group. Determinants of Free Choice What leads some publics to feel that they have free choice? Table 3 shows the factors linked with the rising sense of free choice that was found in most countries surveyed by the Values Surveys from 1981 to Apart from the starting level of people s sense of free choice, three variables predicted the extent to which a country s sense of free choice changed during this period. First, people living in countries that experienced relatively strong economic growth from the earliest to the most recent survey showed a rising sense of free choice. Economic scarcity is one of the most important constraints on the extent to which people actually can make choices, and growing resources enhance freedom of choice. But democratization is of similar importance. The publics of countries that experienced rising levels of democracy also experienced a rising sense of free choice. Indeed, every country that made a transition from authoritarian rule to democracy during this period showed a rising sense of free choice. Even more important than these changes, is the impact of growing tolerance of diverse life styles on people s sense of freedom. As hypothesized, democratization, economic growth and growing social tolerance contributed to a rising feeling that people have free choice and control of their lives. For another perspective on the factors shaping subjective well-being, let us move from OLS regression analysis to a hierarchical linear modeling (HLM) approach that allows us to examine the

18 Development, Freedom, Happiness Perspectives on Psychological Science 18 interactions between societal-level factors and individual-level factors and to identify the microfoundations of the macro-level findings discussed so far. Table 4 shows a series of multi-level regression models using the pooled time series data from 1981 to 2007 on the SWB index. These analyses use the data from all surveys conducted from 1995 to 2007 (N = 206,320). Confirming the findings already reported, economic growth and per capita GDP explained 50% of the country-level differences in SWB (Model 4.1). However, as Sen (2001) argued, the crucial impact of economic development is that it increases freedom of choice. Adding our measure of free choice to the regression, increased the explained variance in SWB from 6% to 15% among individuals and from 50 to 71% among countries, while reducing the impact of the economic variables. A sense of free choice affects people s SWB more as a property of their society than as an individual characteristic: A person s subjective well-being is as more affected by the general atmosphere of freedom in the society in which one lives, than by one s individual sense of freedom. As societies become wealthier, threats to survival become less pressing and people become more tolerant of gender equality and social diversity, and give higher priority to self-expression. An index measuring whether one would accept people of another race, immigrants, or homosexuals as neighbors, shows a significant positive effect at the country-level: people living in more tolerant societies tend to be happier, regardless of their own beliefs (Model 4.3). More open social norms concerning the role of women, ethnic diversity, and alternative lifestyles give people more freedom of choice in how to pursue happiness, and during the past quarter century, tolerance of diversity increased substantially. For example, the proportion of respondents claiming that homosexuality is never justifiable fell from 33% in 1981, to 16% in , in the countries for which data are available from both times. Discriminatory attitudes toward women or racial minorities showed similar downward trends in most countries. As the human development model implies, the impact of free choice on well-being increases as rising incomes increase the utility of freedom. This is evident from the cross-level interaction effects in Model 4.3. The effect of a person s household income on her SWB interacts negatively with the given country s per capita GDP, whereas the effect of a person s sense of free choice on SWB interacts

19 Development, Freedom, Happiness Perspectives on Psychological Science 19 positively with her country s per capita GDP. As societies become wealthier, household income shows a diminishing impact on SWB, but personal freedom shows an increasing impact. When people lack the resources to fulfill their basic needs, the utility of freedom is relatively low: Economic factors are the major determinant of most people s well-being. In more affluent societies, people give higher priority to free choice and self-expression which, accordingly, play an increasingly important role in shaping their well-being. A major reason why people living in economically developed societies have relatively high levels of subjective well-being is because they have a wider range of choice in how to live their lives. Since 1989, dozens of countries have become more democratic; and virtually all high-income and middleincome countries have become more supportive of gender equality and more tolerant of outgroups. In the great majority of the countries for which time series data were available, the proportion of the public saying that they have free choice and control over their lives, increased since According to our interpretation, this implies that we should find rising happiness levels. The strong version of the hedonic treadmill model would dismiss these facts as irrelevant: Any difference between the subjective well-being levels of rich and poor countries must be due to some fixed cultural difference in the meaning of happiness. The low-ranking countries have always been low and will remain so. Insert Figure 3 about here) But they have not. Data from the Values Surveys show that during the past two decades, the SWB index rose in 77% of the countries for which a substantial time series is available. Figure 3 shows the changes observed on the SWB index in all 52 of these countries (with an average of 17 years between the earliest and latest surveys). Contrary to the belief that happiness remains constant, SWB rose in 40 countries and fell in only 12, with a median increase of.35 on this index. Putting it in an intuitively more meaningful form, the average percentage of people in these countries saying they were very happy increased by almost seven points. The probability that these increases are due to chance is negligible. A

20 Development, Freedom, Happiness Perspectives on Psychological Science 20 paired t-test between the earliest and most recent data for countries with at least a ten-year range of data yields a probability of this rise being observed under the null hypothesis of constant global happiness (t = with 42 degrees of freedom). Many of the countries showing the largest increases in SWB are former communist countries. As we will see below, the collapse of communism was generally followed by a sharp decline in SWB, which tended to rise again with economic recovery. In the case of Ukraine and Moldova (the two countries showing the largest increases in SWB), the first available survey was carried out in 1996 overall, a low point following the collapse of communism. The most recent survey from both countries was from 2006, so the strong upward movement may reflect recent recovery, such as the one that occurred in Russia. In Hungary (the country showing the largest decline), the earliest available survey was from 1982 well before the collapse of communism, when SWB was still relatively high and the latest available survey was from We suspect that a more recent survey would show a relative recovery in the Hungarian level of SWB, as was true in Russia. We have substantial time series data for 15 ex-communist countries. In six of them (including Hungary), our most recent survey was from 1999 and five of the six show decreasing life satisfaction. In nine cases, our most recent survey was from , and seven of the nine showed increasing happiness. It seems happiness rose more widely than life satisfaction because life satisfaction is more sensitive to economic conditions than happiness. In most ex-communist countries, the political and social liberation of recent years was accompanied by economic collapse that lasted for about a decade. Only in recent years have their economies recovered. Most of the ex-communist countries for which we do not have recent data show rising happiness but declining life satisfaction (as in Hungary). Most of the excommunist countries for which we do have recent data, show both rising happiness and rising life satisfaction (as in Russia). Happiness and life satisfaction can move in different directions for significant periods of time, but they tend to go together (the societal-level correlation is.81). We suspect that if we had recent data for all of the ex-communist countries, the trend toward rising life satisfaction would be even more widespread.

21 Development, Freedom, Happiness Perspectives on Psychological Science 21 If all six of the ex-communist countries not surveyed since 1999 showed rising life satisfaction, it would raise the percentage of countries with rising life satisfaction to 75%, which is still short of the 88% showing rising happiness. The differential impact of the collapse of communism on happiness and life satisfaction is only part of the explanation for why the former rose more than the latter, but it seems to play an important role. Insert Figure 4 about here We hypothesized that economic development, democratization and increasingly tolerant societies have contributed to a growing sense of freedom and control. In keeping with this interpretation, the public s sense of freedom increased in 79% of the countries for which a substantial time series is available from the Values Surveys. This is an overwhelming trend fully as strong as the global trend toward rising subjective well-being, with which it is closely linked. As the regression analysis in Table 2 demonstrated, a rising sense of free choice is by far the most powerful factor driving rising subjective well-being. By itself, it explains 30% of the changes observed on the SWB index. The fact that, from 1981 to 2007, the publics of most countries experienced a growing sense of free choice seems to be the core reason why subjective well-being has risen. Figure 4 shows how both the sense of freedom and subjective well-being levels increased in most of these countries from the earliest available survey to the latest one and showing that the two have a strong tendency to go together (r =.71). Why Have These Trends Escaped Notice? How did such an important phenomenon as this dramatic rise in subjective well-being escape notice until now? We suggest four reasons. First, most of the earlier evidence came from rich countries that had already passed the point of diminishing returns from economic development, and showed relatively little change. Second, research has largely focused on material factors, despite growing evidence that belief systems and institutions are also important drivers of subjective well-being, and that

22 Development, Freedom, Happiness Perspectives on Psychological Science 22 happiness is shaped by social and psychological factors at least as much as by economic and genetic ones. Third, the decisive social changes global economic growth, widespread democratization, growing tolerance of diversity, and a rising sense of freedom are relatively recent and did not show their impact in earlier surveys. And finally, cross-national research on the determinants of happiness has tended to focus on life satisfaction, rather than happiness per se. This is an important point because the rise in subjective well-being shown in Figure 3 is largely driven by rising happiness levels. While life satisfaction rose in 63% of these societies, happiness increased in 87%. Happiness and life satisfaction are closely correlated, as we have noted, and the two terms are often used interchangeably. Moreover, increases in life satisfaction tended to accompany increases in happiness (r =.33). But although closely linked, they reflect different facets of subjective well-being. Life satisfaction is more strongly tied to a society s economic level, as we saw in Table 1. If one assumes that economic variables are the main determinant of subjective well-being, as much previous research has done, then the fact that life satisfaction is more closely correlated with economic conditions, implies that it is the better measure of SWB. But, as we have seen, economic variables are only part of the story and not the most important part. If instead one views happiness and life satisfaction as tapping complementary aspects of subjective well-being, then one can understand why recent years brought a stronger trend toward rising happiness than toward rising life satisfaction. For while the expansion of freedom was widespread, it was not always accompanied by rising prosperity. In many cases particularly in ex-communist countries democratization was accompanied by economic collapse, leading life satisfaction to fall while happiness increased. Thus, as Figure 5 shows, during the period covered by this analysis, life satisfaction rose in 33 countries and fell in only 19. The rising life satisfaction found in 63% of the countries in our sample constitutes a clear trend toward rising SWB, but it is overshadowed by the trend toward rising happiness shown in Figure 6. Insert Figures 5 and 6 about here